HAL Id: hal-00503999https://hal.archives-ouvertes.fr/hal-00503999

Submitted on 19 Jul 2010

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.

Lab-scale prototype for on-line monitoring and energydiagnosis in buildings

Thierry Talbert, Benjamin Paris, Julien Eynard, Stéphane Grieu, OlivierFruchier

To cite this version:Thierry Talbert, Benjamin Paris, Julien Eynard, Stéphane Grieu, Olivier Fruchier. Lab-scale proto-type for on-line monitoring and energy diagnosis in buildings. International Conference on RenewableEnergy and Eco-Design in Electrical Engineering, Dec 2008, Montpellier, France. pp.0006. �hal-00503999�

Lab-scale prototype for on-line monitoring and energy diagnosis in buildings

Thierry Talbert, Benjamin Paris, Julien Eynard, Stéphane Grieu and Olivier Fruchier

ELIAUS laboratory, University of Perpignan Via Domitia, Perpignan, France

{talbert; benjamin.paris; julien.eynard; grieu; olivier.fruchier}@univ-perp.fr

Introduction

Regarding to greenhouse gases emission or

supplying cost, energy consumption in buildings is

nowadays a problem. So, an appropriate energy

management in buildings is mandatory for enhancing

energy savings and promoting renewable energy

resources. Thus, the present paper focuses on

presenting a lab-scale prototype for (i) on-line

monitoring energy consumption, (ii) on-line

computing a global energy indicator and (iii)

estimating the renewable energy potential of a

location. These three points define an energy

diagnosis based on energy consumption

segmentation. The main goal of developing a lab-

scale prototype is testing some usable tools for

achieving an energy management in buildings by

means of the above-mentioned energy diagnosis.

The lab-scale prototype is composed of a building

mock-up, a monitoring system and a data post-

treatment software. The monitoring system is being

developed, in collaboration with industrial partners,

APEX-BP SOLAR, Pyrescom and the “Centre

Scientifique et Technique du Bâtiment” (CSTB), for

measuring and recording both mock-up

temperatures and energy consumption. In addition,

this monitoring system records also outdoor

parameters, such as solar radiation and wind speed,

for estimating the real renewable energy potential of

the location.

This paper presents, first, the energetic situation in

France and the related documentation, and,

secondly, a detailed presentation of the lab-scale

prototype. A conclusion ends the paper.

Energetic situation in France

Fossil energy consumption is due to different sectors

such as transport or industry. In France, the building

sector represents 42 % of the overall energy

consumed and 25 % of greenhouse gases emissions

[1]. So, the main objective of the French “Grenelle de

l’Environment” is about reducing energy consumption

of 20 % in existing buildings during the five next

years.

French energy documentations

Enhancing the energy performance of buildings, for

promoting energy savings, is already encouraged by

the legal documentation, e.g. the RT 2005

“Réglementation Thermique 2005” [2]. RT 2005

provides the material energy performance to respect

and the conventional energy consumptions for the

future building, taking into account heating,

ventilation, cooling and electricity consumptions.

Conventional energy consumptions also serve of

reference for HPE (“Haute Performance

Energétique”) and THPE (“Très Haute Performance

Energétqiue”) labels, respectively 10 % and 20 % of

consumption reductions.

Moreover, the DPE (“Diagnostic de Performance

Energétique”) [3] was created to characterize the

existing buildings (figure 1), using a global energy

indicator: kWh,m -2,year-1. Some others European

labels exist, PassivHaus and Minergie, using

approximately the same reasoning and the same

indicator.

Fig 1: DPE results.

Software, such as “3CL Excel©” (CSTB), which are

based on building materials parameters (thermal

conductivity insulation, glazing losses …), on the

building design or equipment, can be used to

compute DPE to classify buildings according to their

levels of energy consumption.

The DPE is the first step of an energy diagnosis,

because it is just a “static calculation”. It is more

interesting and real to diagnose the energy

performance of a building in real time. Indeed, to

improve energy management of buildings, we can

consider two complementary approaches:

• optimize the buildings structurally by the

choice of efficient materials,

• use the monitoring tools to control the use

of renewable energy and fossil fuels.

The work presented in this article meet the second

approach to control the energy building behavior.

Energy indicator calculation

Global energy indicator, i.e. kWh,m-2,year-1, used in

the different documentations, can be performed to

enhance energy management in building. It serves of

reference to characterize a building.

The energy indicator calculation is done almost as

follow:

1 register all the energy consumptions of the

building,

2 measured the different energy consumptions

with appropriate sensors,

3 compute the global energy consumption in real

time,

4 add the building area.

From this study, we are able to generate a real

global indicator to define a building. Then, this

buildings could be compared at others buildings,

using the scale of the figure 1, or for example HPE

and THPE labels consumption references (for a

same type of building).

To compute this indicator and work on energy

performance and building classification, a lab-scale

prototype is designed. In addition, the lab-scale

serves to test different controllers on energy facilities

to enhance energy management (see post treatment

software),

Lab-scale prototype

The prototype is divided into two parts:

• a building mock-up, with a complete set of

sensors and two heating resources,

• an electronic device to compute in real time

the mock-up energy indicator, and to control

the heating sources.

Why have-we chosen a mock-up instead of a real

building?

We are looking for a building:

• with several measurements (temperature,

electricity, environmental parameters, ,,,),

• we can manage (whatever the day of the

year and the external temperature),

• where we can easily add sensors, change

command rules etc,

• where two different heating resources are

available (electricity and/or renewable

energy).

Thus, we have to solve 3 problems if we want to

present a real time solution that compute energy

diagnosis:

1 a building with all the sensors. This is possible

but need an human and material investment too

important at this point of the project,

2 a building using two different heating resources

(the buildings of the University of Perpignan use

electricity and fuel for heating but none

renewable energy resources),

3 a building we can control. That's maybe the

bigger problem we have to solve. Changing

regulation to test the monitoring is not possible if

people are there. And more important, we have

to by-pass the regulation integrated into air-

conditioner, hot pump heaters etc. with all the

consequences (destruction of electronic and/or

mechanical, loss of warranty etc).

That's why we have developed this lab-scale

prototype, we'll present on the following paragraphs.

Building mock-up

The prototype of the building is a a mock-up (figure

2), which scale is 1/27, based on a real one floor

house of 128 m2. With this size, the small house

allows fast and simple sensor modifications. This

flexibility provides (i) constructing a simple and

reliable electronic, (ii) choosing the best position for

sensors (iii) the replacement of the heaters.

The mock-up design is done three steps:

1 the physical scale,

2 the materials choice,

3 the instrumentation.

Physical scale

The main idea is the ratio between a real height of

concrete floor (almost 20 cm) and the substitute

material. Indeed, we chose a 6 mm height ceramic

tile for the floor: homogeneity is approximately equal

to concrete. From this ratio, the walls and windows

thickness are directly determined like the building

insulation.

Materials choice

The material choices answers at a few

recommendations:

• the thickness must to respect the predefined

scale,

• the thermal and hygrometry behaviour must

to be more as possible the same as a real

wall.

Thus, we choose 13 mm of plasterboard (BA13) for

Fig 2: Building mock-up.

Sensors

Heaters

the walls, 4 mm of polystyrene for insulation and 4

µm of polyane for the windows.

Instrumentation

Sheath for sensors and actuators cables are located

along the walls and go through the tile. Half of the

roof can be detached in order to modify sensor and

actuators localisation. Eight temperature sensors

composed the mock-up instrumentation: seven

indoor sensors and one outdoor sensor. In addition,

two resistors represent the fossil and the renewable

energy resources, The power of the resistor is

estimated in the following section.

Heating power

To estimate the heating power, several points are

essentials:

1. thermal conductivity for each materials

(W,m -1°C-1),

2. calculated thermal resistivity (m2°C,W -1) using

the material thickness,

3. all the wall areas,

4. calculated losses coefficient Dbât of the mock-up,

including thermal bridges.

The maximum value of the power used to heat the

mock-up is given by an empirical formula (equation

1):

P W =Dbât . S .T sp−T out (eq: 1)

where

– Dbât : losses coefficient W,m -2.°C-1,

– S : heating area m2,

– Tsp : temperature set point °C,

– Tout : reference of the outdoor temperature.

If we chose a temperature set point at 19 °C, and a

reference outdoor temperature of -5 °C, the minimal

value of the heaters is 55 W. This value explains the

two resistors of 22 under 24 VDC inside the mock-

up.

After the building mock-up, the electronic device is

detailed.

Electronic device

The electronic device is realised in collaboration with

industrial partners: Pyrescom SA, Apex-BP Solar

and CSTB. The monitoring station (i) implementation

should be easy and (ii) total cost should remain

rather low. The prototypes of the station is

implemented at three different locations (Apex BP

Solar and Pyrescom headquarters and at the

University of Perpignan). The device presented on

this paper are the laboratory prototypes with heaters

interface (this part is not present on the industrial

version).

The monitoring system is divided into two separated

parts: (i) a core-bloc (composed of a low power

processor, is corresponding memory (RAM and

FLASH), and integrated hosts controllers), and (ii) a

set of adaptable bloc sensors, which means that it is

possible to record and process different data. This

electronic device also uses a Controlled Area

Network (CAN) bus, a low power consumption

processor and an encrypted communication system

to secure data access.

As mentioned, with the chosen architecture, it is

possible to use both, informations concerning energy

consumptions and operating conditions

measurements. The smart transducers transmit data

to the monitoring system, through preferably a CAN

bus.

Figure 4 and 5 shows the electronic device example

of the University’s offices, where one of the

laboratory experimental device is under evaluation.

Fig 4: Prototype of office monitoring.

Fig 3: Laboratory monitoring station without IHM.

Sensors interface

Power supply

Heaters interface

Sensors interface

Core bloc

Core bloc

Researches on heating control systems or on

energetic efficiency showed that north front

temperature and inside temperature measurements

are definitely musts for heating control purposes.

In addition, a compromise was found between the

number of transducers, avoiding information

redundancy and total cost. To generalize this

approach to a broader range of customers, indirect

measurements were preferred whenever possible.

All these sensors can interact with the processor

described.

The core bloc use an ARM9© processor instead of a

micro controller, which is typically used in the

metrology literature, since:

• a low level of energy consumption,

• hosts controllers are already integrated for:

(i) connectivity, (ii) control purposes, (iii)

human interface (CSI, Keypad…), (iv)

memory expansion (MMC, PCMCIA…), and

(v) providing e.g. Bluetooth communication,

• computation power is higher (4-8 bits

versus 32-64 bits, 40 MHz versus 100-400

MHz),

• its high level of memory allows handling a

higher number of different kinds of signals.

• control laws can be implemented, e.g. fault

diagnosis, fuzzy logics, or energy

consumption prediction.

Figure 6 shows how the station is connected to the

mock-up. The core block use a Linux operating

system. So,

• network is active to control and update the

firmware (TCP/IP protocols), no need to

program this communication feature.

• security (i,e, encryption) protocols are

enabled. Controlling, uploading,

downloading files (measurements done by

sensors, parameters for sampling and for

scaling sensors etc.) are done on a secured

environment.

• a web server is active. It uses a small

memory footprint, thus following activity of

the station (measured data, error detection,

sensor failure etc.) is possible even if we are

not directly connected to it.

Fig 6: Monitoring station input/output and mock-up connections.

In parallel, the electricity consumption measurement

of the monitoring system must be performed to be

complete concerning the energy management.

Results are shown on Figure 7,

We can establish three working parts:

• Part A: This is the boot of the system. The

loader of the station is waiting 30s for a

command (like uploading a new firmware,

changing boot parameters etc.) and if there

is no actions, the boot of the operating

system is started. At this point all the

components are powered-up but with no

activity, except the processor ARM9, the

RAM and the RS232 connector. Power

consumption is only 1,33 W.

• Part B: This is the boot of the operating

system. It operates first by running the

Linux kernel (section B1), and next loading

all the necessary software (web, acquisition,

secure access etc.) in section B2. On the

first stage, access are only between the

processor and the RAM. On the second

stage, this is between the processor, the

RAM and the FLASH where is stored the

OS. Thus, the power consumption is higher

than on the previous stage: 0,28 W more

important.

• Part C: here the processor is 90% of is time

idle. The 10 % remaining, is recovering data

from sensors, compute indicators, save

data on the FLASH and verify heaters

command. So, power consumption is

almost the same as section A of the figure

(1,33 W). The 3 spots (C1, C2 and C3)

Fig 5: Smart transducers and monitoring system localization (University Offices). Power interface

Core blocARM9 + DAC + ADC

Sensorsinterface

RJ45Serveurs

ssh, sftp, web

RS232 USB

Actuators(heaters, air-conditionners etc.)

Sensors(temperature, humidity etc.)

Mock-up building

represented on figure 7, show when the

processor retrieve data from the sensors

(C1 and C3) and when the data is stored

(C2). Power consumption is higher

because, the CPU work at 100 % and all the

communication channels are active

(FLASH/CPU, CPU/RAM, CPU/SPI,

CPU/CAN bus etc.).

With this structure, we can complete the monitoring

system with actuators and with optimal regulation to

increase energy building efficiency.

On-line measurements

Figure 8 and 9 focus on the lab-scale on-line

measurements using the electronic device, Several

heating powers are send to the electronic device, and

all the mock-up temperatures are recorded,

Fig 8: Heating powers.

Fig 9: Mock-up temperatures.

Furthermore on figure 10 and 11, on-line

measurements and heaters status are available on a

web site (loaded all the five minutes).

Fig 10: Real-time temperature measurements done inside and outside the mock-up. (Values are chown

in mV, and scaled in °C by the station main program.)

Fig 7: Boot consumption of the station.

A B C

190mA

230mA

70mA C1 C3

C2

B1 B2

Post-treatment software

The post-treatment software proposes three main

works:

• the computation of the global energy

indicator already explained

• the use of advance heating controllers

• the calculation of different renewable energy

potentials of a location.

First, some alterations to the monitoring software

allows adapting the calculation of the global energy

indicator to European labels such as the German

PassivHaus or Swiss Minergie labels. Indeed, each

labels takes into account different energy

consumptions (heating, cooling, ventilation,,,,) and

several parameters, The monitoring device is

adaptable to gather all the different data required,

Secondly, different advance heating controllers [4]

such as PID, MPC, fuzzy logic..., can also be added

for enhancing the use of renewable energy resources

and the energy savings. The post-treatment software

controls for example the mean temperature of a

room, with several temperature set points.

In the end, the post-treatment software determines,

after an installation of the electronic device into a

building, the real potential of the location concerning

solar (figure 12) and wind (figure 13) renewable

energies. A set of sensors composed by a spekton

300 (solar radiation), an anemometer and a

weathercock (made at the laboratory) are connected

to the device (figure 14).

The solar and wind speed measurements allow

proposing renewable energy resources integration (in

this case photovoltaic, solar thermal and wind

turbine). Cost of the modification(s) is evaluated and

the return of investment (ROI) also.

In addition, with the energy diagnosis, an evaluation

of buildings modifications, to increase efficiency

(including material changes, new sources integration

etc.) can be realized.

Conclusions

The aim of the present paper is presenting a plug

and play electronic device, connected to a full set of

sensors, for evaluating in real-time the energetic

efficiency of buildings. This low-cost and low-

Fig 11: On-line measurement of.a heating power.

Fig 12: Anemometer and waethercock sensors on the roof of the University.

Fig 13: Solar radiation measured the 2008/07/01.

Fig 14: Wind speed measuremrnt done the 2008/07/01.

consumption device allows measuring energy

consumptions and computing an energetic indicator

for carrying out an energy diagnosis. This diagnosis

is done during a data post-treatment phase which

provides the real renewable energy potential and

allows controlling energy facilities [5]. A building

mock-up has been designed for testing, in real

running conditions, the above-mentioned electronic

device and its applications.

A commercial version of both electronic device and

post-treatment software is under development by our

above-mentioned industrial partners.

Future work will now focus on implementing the

developed electronic device on real buildings,

starting from industrial partner's offices.

Acknowledgements

The authors acknowledges the financial support

received from FCE (Fonds de Compétitivité des

Entreprises). We want also to thank our partners, the

enterprises: Pyrescom SA, Apex-BP Solar and CSTB

(Centre Scientifique et Technique du Bâtiment).

Without them, this work would not have been

possible.

References

1 ADEME (Agence De l’Environnement et de la

Maîtrise de l’Energie), « Les chiffres clés du bâtiment

en 2006 », Publications de l’ADEME, France, 2007

2 République française, « Décret n°2006-592 du 24

mai 2006 relatif aux caractéristiques thermiques et à

la performance énergétique des constructions »,

Journal Officiel, France, 2006.

3 République Française, « Décret n°2006-11-47 du

14 septembre 2006 relatif au diagnostic de

performance énergétique et à l’état de l’installation

intérieure », Journal Officiel, France, 2006.

4 B. Paris, J. Eynard, G. François, T. Talbert and M.

Polit, « A prototype for on-line monitoring and

control of energy performance for renewable energy

buildings », ICINCO 2008, 5th International

Conference on Informatics in Control, Automation &

Robotics, Funchal, Portugal, pp. 125:130, 2008.

5 B. Paris, J. Eynard, F. Thiery, A. Traoré, T.

Talbert and S. Grieu « Fuzzy-PID control for

multisource energy management in buildings »,

IREED 2008, Montpellier.